We first repeat some well known facts about condition numbers for normwise and componentwise perturbations. Consider the matrix

Size: px
Start display at page:

Download "We first repeat some well known facts about condition numbers for normwise and componentwise perturbations. Consider the matrix"

Transcription

1 BIT 39(1), pp , 1999 ILL-CONDITIONEDNESS NEEDS NOT BE COMPONENTWISE NEAR TO ILL-POSEDNESS FOR LEAST SQUARES PROBLEMS SIEGFRIED M. RUMP Abstract. The condition number of a problem measures the sensitivity of the answer to small changes in the input, where small refers to some distance measure. A problem is called illconditioned if the condition number is large, and it is called ill-posed if the condition number is infinity. It is known that for many problems the (normwise) distance to the nearest ill-posed problem is proportional to the reciprocal of the condition number. Recently it has been shown that for linear systems and matrix inversion this is also true for componentwise distances. In this note we show that this is no longer true for least squares problems and other problems involving rectangular matrices. Problems are identified which are arbitrarily ill-conditioned (in a componentwise sense) whereas any componentwise relative perturbation less than 1 cannot produce an ill-posed problem. Bounds are given using additional information on the matrix. Key words. Componentwise distance, condition number, ill-posed, least squares, pseudoinverse, underdetermined linear systems AMS subject classifications. 65F20, 65G05 We first repeat some well known facts about condition numbers for normwise and componentwise perturbations. Consider the matrix A = A(e) := 1 e for 0 e R. 1 0 e For small values of e, a small (normwise) change of the matrix components produces a singular matrix. Indeed, A + is singular for = 0 e , and 2 = 0 0 e = 1 = e. However, the matrix represents a large componentwise relative perturbation of the entries A 12 and A 33, and expanding the determinant of A by Cramer s rule it is clear that any componentwise relative perturbation of A less than 1 cannot produce a singular matrix. The problem of matrix inversion of A is ill-conditioned in a normwise sense because small normwise perturbations of A may produce large normwise changes of A 1. Likewise, the problem of matrix inversion of A is well-conditioned subject to componentwise relative perturbations of A. This is because a small componentwise relative perturbation of A causes a small relative change of the determinant of A (as the sum of two numbers of the same sign), and the determinant of all 2 2 submatrices of A is equal to the product of two (sub)matrix entries. What is true in this specific example turned out to be a general fact: If the problem of matrix inversion is ill-conditioned, then the distance to the nearest singular matrix is small using the same distance measure, no matter whether normwise or componentwise distances are used. In this note we will investigate whether this can be extended to other problems. Consider for a real m n matrix A and a real vector b with m components the following five problems (a pseudoinverse always refers to the Moore-Penrose or (1,2,3,4) Inst. f. Informatik III, Technical University Hamburg-Harburg, Eißendorfer Str. 38, Hamburg, Germany (rump@tu-harburg.de). 1

2 inverse): i) A system of linear equations Ax = b for m = n, ii) matrix inversion A 1 for m = n, iii) computation of the pseudoinverse A + for m n, iv) a least squares problem Ax b 2 = Min! for m > n, v) an underdetermined linear system x 2 = Min! subject to Ax b 2 = Min! for m < n. For a small perturbation of the data the change in the solution of one of these problems can be estimated by means of the condition number. Define for a nonsingular square matrix A, (1) κ(a) := A 1 A. For simplicity we use throughout this note the spectral norm. We mention that most of the following results are valid for other norms as well. Consider a perturbation à of A with à A ε A. If, for any right hand side, x is the solution of the perturbed system à x = b, then [11, Theorem III.2.11] (2) x x x ε κ(a), and for nonsingular à [11, Corollary III.2.7] (3) à 1 A 1 ε κ(a). à 1 When replacing x, à 1 by the true solution x, A 1 in the denominator of (2), (3), respectively, the right hand sides is to be replaced by ε κ(a)/(1 ε κ(a)). Moreover, these bounds are attainable up to a small factor (see [5, Theorem 7.2], [11, Corollary III.2.7] and the discussion over there). This gave reason to call κ(a) the condition number of the matrix (rather than of a problem). The bounds (2) and (3) look very similar for the other three problems involving rectangular matrices. The definition of the condition number (1) is generalized to a rectangular matrix of full rank into (4) κ(a) := A + A. Then, for a perturbed matrix à with à A ε A and rank(ã) = rank(a), it is [11, Corollary III.3.10] Ã+ A + ε 2 κ(a). Ã+ Similarly, for ε κ(a) < 1 and any right hand side, the solution x of the original problem and the solution x of the perturbed problem with r = Ax b, r = à x b, satisfy [4, Theorem 5.3.1] r r b ε (1 + 2 κ(a)) + O(ε 2 ) 2

3 for least squares problems, and [4, Theorem 5.7.1] x x x ε 2 κ(a) + O(ε 2 ) for underdetermined linear systems. Therefore κ(a) := A + A is often called the condition number of a (full-rank) rectangular matrix. Note that the sensitivity of the solution (rather than of the residual) of least squares problems is the square of κ(a) [4, Theorem 5.3.1]. It is well known that any of the above five problems becomes ill-posed [12] if and only if the matrix is rank-deficient. The 2-norm distance of a full-rank matrix to the nearest singular matrix is the smallest singular value of A [4, Theorem 2.5.2], which is A + 1. Therefore, for full-rank square and rectangular matrices, (5) min { 0 < ε R This confirms the well known fact that (6) à : à A ε A, à rank-deficient the normwise distance to the nearest ill-posed problem of a problem i) to v) is proportional to the reciprocal of κ(a). } = κ(a) 1. This has been shown to be true for a number of other problems in numerical analysis in the classical paper [2]. It seems natural to ask whether this statement is still true for other measures of distance, for example componentwise distances. For this purpose we need an appropriate condition number. We restrict our attention to the practically important case of componentwise relative perturbations but mention that most of the following is valid in a much more general setting, i.e. for componentwise perturbations subject to an arbitrary nonnegative weight matrix [8,?]. For a given matrix A (square or rectangular) consider a perturbation à with (7) à A ε A. Here and in the following, absolute value and comparison are always to be understood componentwise. So (7) is equivalent to Ãij A ij ε A ij for all i, j. Define for a full-rank matrix A, (8) cond(a) := A + A. For square matrices this is the so-called Bauer-Skeel condition number. Let à be a perturbation of A satisfying (7). Then for a square matrix A and any right hand side b with Ax = b and à x = b and ε cond(a) < 1, it is [11, Corollary III.2.15] x x x ε cond(a) 1 ε cond(a). For a square or underdetermined linear system with full-rank matrix A and ε cond(a) < 1, it is [5, (20.8)] x x x ε 3 cond(a) + O(ε 2 ). There are similar, although more involved, estimates for the sensitivity of matrix inversion and least squares problems subject to componentwise relative perturbations 3

4 (see, for example, [5, Theorem 7.4 and Theorem 19.2]). These estimations contain again cond(a). Therefore, cond(a) := A + A is sometimes called the componentwise condition number of a matrix. The question is whether some relation similar to (5) can be established in the case of componentwise distances. Define for a full-rank rectangular matrix A, (9) σ(a, A ) := min { 0 < α R Ã : Ã A α A, Ã rank-deficient Definition (9) can be generalized to other nonnegative weight matrices E by replacing A by E. In the following, we restrict our attention to the important case of entrywise relative perturbations: E = A. Obviously 0 < σ(a, A ) 1. Unlike in the normwise case, a characterization cannot be a simple equality as in (5). This is because Poljak and Rohn showed that computation of σ(a, A ) for square A is NP-hard [7]. Moreover, the definition of cond(a) still depends on a norm, and improper scaling may produce artificial illconditioning [11, p.122f]. Let us first consider the square case, i.e. problems i) and ii). In this case we may ask for the optimal scaling, that is for the minimum achievable condition number cond(ad) for nonsingular diagonal D (the condition number cond(a) is independent of row scaling for m n). For ρ denoting the spectral radius it is ([3, 5], see also [1]) }. (10) inf D cond(ad) = ρ( A 1 A ), where for irreducible A 1 A the minimizing diagonal matrix can be computed explicitly from the left and right Perron vector of A 1 A. The question is whether, similar to the normwise case (5), the minimum achievable componentwise condition number is related to the componentwise distance to the nearest rank-deficient matrix. Indeed, it has been shown in [?, 8] that for square nonsingular A (11) inf D 1 σ(a, A ) cond(ad) ( )n inf cond(ad). D The left bound has been known for long and is an immediate consequence of Perron- Frobenius theory: If A + = A(I + A 1 ) is singular for β A and β := σ(a, A ), then 1 ρ(a 1 ) ρ( A 1 ) ρ( A 1 ) β ρ( A 1 A ). The right bound of (11) is the difficult part; a consolidated proof of this inequality will appear in SIAM Review [10]. The right bound is sharp up to the factor for any n [?]. In words: (12) The componentwise distance to the nearest rank-deficient matrix is proportional to the reciprocal of the minimum achievable componentwise condition number for square matrices. That means, if a linear system or the problem of matrix inversion is ill-conditioned subject to componentwise relative perturbations of the matrix, then a small componentwise relative perturbation produces an ill-posed problem. It seems natural to ask whether this statement is still true for problems iii) to v). First, we face the same problem as in the square case that cond(a) depends on a norm. For full-rank least squares problems it is (13) inf D cond(ad) = ρ( A+ A ) for A M mn (R), m n. 4

5 This is because (AD) + = D 1 A +, and setting B := A + A, (13) is equivalent to inf D D 1 BD = ρ(b) for B M n (R), B 0, which can be shown along the lines of the proof of Theorem 5.2 in [3]. Unlike the corresponding condition number for square matrices, κ(a) is not independent of diagonal row scaling. However, it does not make sense to use an infimum of κ(da) over diagonal D, because this changes the problem, and putting zeros in appropriate diagonal positions of D readily produces the (usual) condition number of some square matrix out of A. Furthermore, for a least squares problem row scaling is usually related to uncertainties in the right hand side (cf. [6]). The same remarks apply similarly to undetermined linear systems. In any case, for square as well as for rectangular matrices, it is (14) ρ( A + A ) cond(a) and ρ( A + A ) A + A mn κ(a). If a statement in the spirit of (12) would be true for one of the problems iii) to v), then for a rectangular matrix A with ρ( A + A ) being large, and henceforth cond(a) being large, a small componentwise relative perturbation must produce a rank-deficient matrix. In the following we show that this is not true in general. Suppose A is a real m n matrix of full rank k = min(m, n), and à is some rank-deficient perturbation of A. The rank-deficiency of à is equivalent to the fact that all k k submatrices of à are singular. Therefore, definition (9) yields for a real m n matrix A of full rank k = min(m, n) (15) max A σ(a, A ) σ(a, A ) min A σ(a, A ), where the maximum and minimum is taken over all k k submatrices A of A. Consider the following matrix (16) A = A(ε) := ε for 0 ε R. Then det(a T A) = 6ε 2 and A is of full rank. Furthermore, for ε < 2/3, A + A = ε 2 ε 3 ε 2 ε 3 ε ε implies ρ( A + A ) > ε 1. By (14), κ(a) is large for small ε, and by (6) a small normwise perturbation must produce a rank-deficient matrix. Indeed, changing A 33 into 0 yields a rank-2 matrix. All three problems iii) to v) are also ill-conditioned subject to small relative perturbations of A. It is (the numbers were computed using Maple [13]) A + = 1 6ε 2 2 3ε 6 2 3ε 2ε 2ε 0 2ε ,

6 whereas perturbing A into à by changing the (4,3)-component into 1 δ yields (Ã+ ) 11 = 1 ( ( ) ) 10 6ε 2 + 3ε 1 δ + O(δ 2 ). Therefore, the sensitivity of (Ã+ ) 11 with respect to a componentwise relative perturbation of A 43 is larger than ε 1. Thus the problem of computing the pseudoinverse of A, the least squares problem with right hand side (1, 0, 0) T, and the underdetermined linear system with right hand side (1, 0, 0, 0) T is indeed ill-conditioned for componentwise relative perturbations. On the other hand, we know from (15) that the componentwise distance to the nearest rank-deficient matrix is bounded below by the componentwise distance of the upper 3 3 matrix of A to the nearest singular matrix. The determinant of this matrix is A 12 A 21 A 33, and any componentwise relative perturbation less than 1 of A cannot produce a singular matrix. Summarizing: If the problem of computing the pseudoinverse, a least squares problem, or an underdetermined linear system is ill-conditioned subject to componentwise relative perturbations of the matrix entries, then the problem need not be near an ill-posed problem subject to componentwise relative perturbations. In fact, in our example the componentwise distance to the nearest rank-deficient matrix is equal to the distance to the rank-0 matrix. From the 4 3 example (16) it is easy to derive examples of larger dimensions. The main difference between the square and the rectangular case is that, for k = min(m, n), in the latter case a perturbation has to lower the rank of every k k submatrix simultaneously, whereas in the first case only one matrix, the given matrix, must become singular. Another explanation uses that for à := A+ the matrix à T à = (A + ) T (A + ) must become singular for a perturbation. For a similar reason, a symmetric matrix, the inversion of which is sensitive with respect to symmetric componentwise relative perturbations, need not be near a singular matrix with respect to symmetric componentwise relative perturbations [9]. For the square cases i) and ii) we have with (11) general two-sided bounds for the ratio between the componentwise distance to the nearest ill-posed problem and the reciprocal of the condition number. Example (16) shows that this ratio is, in general, unbounded for the problems iii) to v). The question remains whether bounds are possible using additional information on A. We are interested in bounds for the quantity γ := σ(a, A ) cond(a) = σ(a, A ) A + A for full-rank rectangular A M mn (R). Note that, unlike (11), we use the actual condition number, not an optimal condition number subject to some scaling. Such bounds are indeed possible. A general lower bound without further assumptions on A can be established similar to the square case. Let full-rank A M mn (R), m n, and rank deficient 6

7 A + = (I + A + )A with σ(a, A ) A be given. Then I + A + is rankdeficient and (17) 1 ρ( A + ) = ρ(a + ) ρ( A + ) σ(a, A ) ρ( A + A ) = γ. For m < n the same argument using A + = A(I + A + ) applies. For an upper bound we need additional information on A. Denote µ := min A ij and M := max A ij. i,j i,j For the singular value decomposition A = σ i u i v T i with smallest singular value σ k, k = min(m, n), the matrix A σ k u k v T k is rank-deficient. For (1) mn denoting the m n matrix of all 1 s, σ k u k v T k σ k (1) mn µ 1 σ k A = µ 1 A + 1 A. Hence σ(a, A ) {µ A + } 1. On the other hand [5, Table 6.2], cond(a) = A + A A + A 1 A + A A + 1 A 1 A + A M mn A + 1 A + M(mn) 3/4 A +. Together with (17) and (16) this proves the following result for square and for rectangular matrices. For square matrices a weaker result of this kind is [3, Theorem 5.4]. Theorem 1.1. Suppose A M mn (R) has full rank, suppose A ij 0 for all i, j, and define Then ϕ := max A ij min A ij. 1 σ(a, A ) ϕ cond(a) (mn)3/4 cond(a), where the componentwise condition number cond(a) is defined by (8), and the componentwise distance σ(a, A ) to the nearest rank-deficient problem subject to relative componentwise perturbations is defined by (9). Without the assumption A ij 0, the product σ(a, A ) cond(a) can be arbitrarily large for m n, m 4, n 3. The result can be interpreted as follows. The closer the entries of A are in magnitude, that means the more a componentwise relative perturbation becomes a componentwise absolute perturbation, the closer are the componentwise distance to the nearest rank-deficient matrix and the reciprocal of the condition number. Acknowledgement. The author is indebted to Nick Higham and two anonymous referees for their thorough reading, constructive comments and valuable suggestions improving the text and exposition of the paper. 7

8 REFERENCES [1] F.L. Bauer. Optimally scaled matrices. Numerische Mathematik 5, pages 73 87, [2] J.B. Demmel. Condition Numbers and the Distance to the Nearest Ill-posed Problem. Numer. Math. 51, pages , [3] J.W. Demmel. The Componentwise Distance to the Nearest Singular Matrix. SIAM J. Matrix Anal. Appl., 13(1):10 19, [4] G.H. Golub and Ch. Van Loan. Matrix Computations. Johns Hopkins University Press, second edition, [5] N.J. Higham. Accuracy and Stability of Numerical Algorithms. SIAM Publications, Philadelphia, [6] C.L. Lawson and R.J. Hanson. Solving Least Squares Problems. Prentice-Hall, [7] S. Poljak and J. Rohn. Checking Robust Nonsingularity Is NP-Hard. Math. of Control, Signals, and Systems 6, pages 1 9, [8] S.M. Rump. Verified Solution of Large Systems and Global Optimization Problems. J. Comput. Appl. Math., 60: , [9] S.M. Rump. Structured Perturbations and Symmetric Matrices. Linear Algebra and its Applications (LAA), 278: , [10] S.M. Rump. The sign-real spectral radius and cycle products. Linear Algebra and its Applications (LAA), 279: , [11] G.W. Stewart and J. Sun. Matrix Perturbation Theory. Academic Press, [12] A.N. Tikhonov and V.L. Arsenin. Solutions of Ill-posed Problems. John Wiley, New York, [13] Maple V. Release 5.0, Reference Manual,

For δa E, this motivates the definition of the Bauer-Skeel condition number ([2], [3], [14], [15])

For δa E, this motivates the definition of the Bauer-Skeel condition number ([2], [3], [14], [15]) LAA 278, pp.2-32, 998 STRUCTURED PERTURBATIONS AND SYMMETRIC MATRICES SIEGFRIED M. RUMP Abstract. For a given n by n matrix the ratio between the componentwise distance to the nearest singular matrix and

More information

Componentwise perturbation analysis for matrix inversion or the solution of linear systems leads to the Bauer-Skeel condition number ([2], [13])

Componentwise perturbation analysis for matrix inversion or the solution of linear systems leads to the Bauer-Skeel condition number ([2], [13]) SIAM Review 4():02 2, 999 ILL-CONDITIONED MATRICES ARE COMPONENTWISE NEAR TO SINGULARITY SIEGFRIED M. RUMP Abstract. For a square matrix normed to, the normwise distance to singularity is well known to

More information

OPTIMAL SCALING FOR P -NORMS AND COMPONENTWISE DISTANCE TO SINGULARITY

OPTIMAL SCALING FOR P -NORMS AND COMPONENTWISE DISTANCE TO SINGULARITY published in IMA Journal of Numerical Analysis (IMAJNA), Vol. 23, 1-9, 23. OPTIMAL SCALING FOR P -NORMS AND COMPONENTWISE DISTANCE TO SINGULARITY SIEGFRIED M. RUMP Abstract. In this note we give lower

More information

Interval solutions for interval algebraic equations

Interval solutions for interval algebraic equations Mathematics and Computers in Simulation 66 (2004) 207 217 Interval solutions for interval algebraic equations B.T. Polyak, S.A. Nazin Institute of Control Sciences, Russian Academy of Sciences, 65 Profsoyuznaya

More information

On the Perturbation of the Q-factor of the QR Factorization

On the Perturbation of the Q-factor of the QR Factorization NUMERICAL LINEAR ALGEBRA WITH APPLICATIONS Numer. Linear Algebra Appl. ; :1 6 [Version: /9/18 v1.] On the Perturbation of the Q-factor of the QR Factorization X.-W. Chang McGill University, School of Comptuer

More information

Subset selection for matrices

Subset selection for matrices Linear Algebra its Applications 422 (2007) 349 359 www.elsevier.com/locate/laa Subset selection for matrices F.R. de Hoog a, R.M.M. Mattheij b, a CSIRO Mathematical Information Sciences, P.O. ox 664, Canberra,

More information

BOUNDS OF MODULUS OF EIGENVALUES BASED ON STEIN EQUATION

BOUNDS OF MODULUS OF EIGENVALUES BASED ON STEIN EQUATION K Y BERNETIKA VOLUM E 46 ( 2010), NUMBER 4, P AGES 655 664 BOUNDS OF MODULUS OF EIGENVALUES BASED ON STEIN EQUATION Guang-Da Hu and Qiao Zhu This paper is concerned with bounds of eigenvalues of a complex

More information

On the Skeel condition number, growth factor and pivoting strategies for Gaussian elimination

On the Skeel condition number, growth factor and pivoting strategies for Gaussian elimination On the Skeel condition number, growth factor and pivoting strategies for Gaussian elimination J.M. Peña 1 Introduction Gaussian elimination (GE) with a given pivoting strategy, for nonsingular matrices

More information

On the Solution of Constrained and Weighted Linear Least Squares Problems

On the Solution of Constrained and Weighted Linear Least Squares Problems International Mathematical Forum, 1, 2006, no. 22, 1067-1076 On the Solution of Constrained and Weighted Linear Least Squares Problems Mohammedi R. Abdel-Aziz 1 Department of Mathematics and Computer Science

More information

WHEN MODIFIED GRAM-SCHMIDT GENERATES A WELL-CONDITIONED SET OF VECTORS

WHEN MODIFIED GRAM-SCHMIDT GENERATES A WELL-CONDITIONED SET OF VECTORS IMA Journal of Numerical Analysis (2002) 22, 1-8 WHEN MODIFIED GRAM-SCHMIDT GENERATES A WELL-CONDITIONED SET OF VECTORS L. Giraud and J. Langou Cerfacs, 42 Avenue Gaspard Coriolis, 31057 Toulouse Cedex

More information

Review of Some Concepts from Linear Algebra: Part 2

Review of Some Concepts from Linear Algebra: Part 2 Review of Some Concepts from Linear Algebra: Part 2 Department of Mathematics Boise State University January 16, 2019 Math 566 Linear Algebra Review: Part 2 January 16, 2019 1 / 22 Vector spaces A set

More information

THE PERTURBATION BOUND FOR THE SPECTRAL RADIUS OF A NON-NEGATIVE TENSOR

THE PERTURBATION BOUND FOR THE SPECTRAL RADIUS OF A NON-NEGATIVE TENSOR THE PERTURBATION BOUND FOR THE SPECTRAL RADIUS OF A NON-NEGATIVE TENSOR WEN LI AND MICHAEL K. NG Abstract. In this paper, we study the perturbation bound for the spectral radius of an m th - order n-dimensional

More information

A Note on Eigenvalues of Perturbed Hermitian Matrices

A Note on Eigenvalues of Perturbed Hermitian Matrices A Note on Eigenvalues of Perturbed Hermitian Matrices Chi-Kwong Li Ren-Cang Li July 2004 Let ( H1 E A = E H 2 Abstract and à = ( H1 H 2 be Hermitian matrices with eigenvalues λ 1 λ k and λ 1 λ k, respectively.

More information

A MODIFIED TSVD METHOD FOR DISCRETE ILL-POSED PROBLEMS

A MODIFIED TSVD METHOD FOR DISCRETE ILL-POSED PROBLEMS A MODIFIED TSVD METHOD FOR DISCRETE ILL-POSED PROBLEMS SILVIA NOSCHESE AND LOTHAR REICHEL Abstract. Truncated singular value decomposition (TSVD) is a popular method for solving linear discrete ill-posed

More information

arxiv: v1 [math.na] 1 Sep 2018

arxiv: v1 [math.na] 1 Sep 2018 On the perturbation of an L -orthogonal projection Xuefeng Xu arxiv:18090000v1 [mathna] 1 Sep 018 September 5 018 Abstract The L -orthogonal projection is an important mathematical tool in scientific computing

More information

Permutation transformations of tensors with an application

Permutation transformations of tensors with an application DOI 10.1186/s40064-016-3720-1 RESEARCH Open Access Permutation transformations of tensors with an application Yao Tang Li *, Zheng Bo Li, Qi Long Liu and Qiong Liu *Correspondence: liyaotang@ynu.edu.cn

More information

A Backward Stable Hyperbolic QR Factorization Method for Solving Indefinite Least Squares Problem

A Backward Stable Hyperbolic QR Factorization Method for Solving Indefinite Least Squares Problem A Backward Stable Hyperbolic QR Factorization Method for Solving Indefinite Least Suares Problem Hongguo Xu Dedicated to Professor Erxiong Jiang on the occasion of his 7th birthday. Abstract We present

More information

Review Questions REVIEW QUESTIONS 71

Review Questions REVIEW QUESTIONS 71 REVIEW QUESTIONS 71 MATLAB, is [42]. For a comprehensive treatment of error analysis and perturbation theory for linear systems and many other problems in linear algebra, see [126, 241]. An overview of

More information

Modified Gauss Seidel type methods and Jacobi type methods for Z-matrices

Modified Gauss Seidel type methods and Jacobi type methods for Z-matrices Linear Algebra and its Applications 7 (2) 227 24 www.elsevier.com/locate/laa Modified Gauss Seidel type methods and Jacobi type methods for Z-matrices Wen Li a,, Weiwei Sun b a Department of Mathematics,

More information

NP-hardness results for linear algebraic problems with interval data

NP-hardness results for linear algebraic problems with interval data 1 NP-hardness results for linear algebraic problems with interval data Dedicated to my father, Mr. Robert Rohn, in memoriam J. Rohn a a Faculty of Mathematics and Physics, Charles University, Malostranské

More information

Section 3.9. Matrix Norm

Section 3.9. Matrix Norm 3.9. Matrix Norm 1 Section 3.9. Matrix Norm Note. We define several matrix norms, some similar to vector norms and some reflecting how multiplication by a matrix affects the norm of a vector. We use matrix

More information

arxiv: v1 [cs.na] 7 Jul 2017

arxiv: v1 [cs.na] 7 Jul 2017 AE regularity of interval matrices Milan Hladík October 4, 2018 arxiv:1707.02102v1 [cs.na] 7 Jul 2017 Abstract Consider a linear system of equations with interval coefficients, and each interval coefficient

More information

STAT 309: MATHEMATICAL COMPUTATIONS I FALL 2017 LECTURE 5

STAT 309: MATHEMATICAL COMPUTATIONS I FALL 2017 LECTURE 5 STAT 39: MATHEMATICAL COMPUTATIONS I FALL 17 LECTURE 5 1 existence of svd Theorem 1 (Existence of SVD) Every matrix has a singular value decomposition (condensed version) Proof Let A C m n and for simplicity

More information

Multiplicative Perturbation Bounds of the Group Inverse and Oblique Projection

Multiplicative Perturbation Bounds of the Group Inverse and Oblique Projection Filomat 30: 06, 37 375 DOI 0.98/FIL67M Published by Faculty of Sciences Mathematics, University of Niš, Serbia Available at: http://www.pmf.ni.ac.rs/filomat Multiplicative Perturbation Bounds of the Group

More information

LinGloss. A glossary of linear algebra

LinGloss. A glossary of linear algebra LinGloss A glossary of linear algebra Contents: Decompositions Types of Matrices Theorems Other objects? Quasi-triangular A matrix A is quasi-triangular iff it is a triangular matrix except its diagonal

More information

Structured Condition Numbers of Symmetric Algebraic Riccati Equations

Structured Condition Numbers of Symmetric Algebraic Riccati Equations Proceedings of the 2 nd International Conference of Control Dynamic Systems and Robotics Ottawa Ontario Canada May 7-8 2015 Paper No. 183 Structured Condition Numbers of Symmetric Algebraic Riccati Equations

More information

Weaker assumptions for convergence of extended block Kaczmarz and Jacobi projection algorithms

Weaker assumptions for convergence of extended block Kaczmarz and Jacobi projection algorithms DOI: 10.1515/auom-2017-0004 An. Şt. Univ. Ovidius Constanţa Vol. 25(1),2017, 49 60 Weaker assumptions for convergence of extended block Kaczmarz and Jacobi projection algorithms Doina Carp, Ioana Pomparău,

More information

On The Belonging Of A Perturbed Vector To A Subspace From A Numerical View Point

On The Belonging Of A Perturbed Vector To A Subspace From A Numerical View Point Applied Mathematics E-Notes, 7(007), 65-70 c ISSN 1607-510 Available free at mirror sites of http://www.math.nthu.edu.tw/ amen/ On The Belonging Of A Perturbed Vector To A Subspace From A Numerical View

More information

Some bounds for the spectral radius of the Hadamard product of matrices

Some bounds for the spectral radius of the Hadamard product of matrices Some bounds for the spectral radius of the Hadamard product of matrices Guang-Hui Cheng, Xiao-Yu Cheng, Ting-Zhu Huang, Tin-Yau Tam. June 1, 2004 Abstract Some bounds for the spectral radius of the Hadamard

More information

Diagonal and Monomial Solutions of the Matrix Equation AXB = C

Diagonal and Monomial Solutions of the Matrix Equation AXB = C Iranian Journal of Mathematical Sciences and Informatics Vol. 9, No. 1 (2014), pp 31-42 Diagonal and Monomial Solutions of the Matrix Equation AXB = C Massoud Aman Department of Mathematics, Faculty of

More information

On condition numbers for the canonical generalized polar decompostion of real matrices

On condition numbers for the canonical generalized polar decompostion of real matrices Electronic Journal of Linear Algebra Volume 26 Volume 26 (2013) Article 57 2013 On condition numbers for the canonical generalized polar decompostion of real matrices Ze-Jia Xie xiezejia2012@gmail.com

More information

Linear Algebra Massoud Malek

Linear Algebra Massoud Malek CSUEB Linear Algebra Massoud Malek Inner Product and Normed Space In all that follows, the n n identity matrix is denoted by I n, the n n zero matrix by Z n, and the zero vector by θ n An inner product

More information

Yimin Wei a,b,,1, Xiezhang Li c,2, Fanbin Bu d, Fuzhen Zhang e. Abstract

Yimin Wei a,b,,1, Xiezhang Li c,2, Fanbin Bu d, Fuzhen Zhang e. Abstract Linear Algebra and its Applications 49 (006) 765 77 wwwelseviercom/locate/laa Relative perturbation bounds for the eigenvalues of diagonalizable and singular matrices Application of perturbation theory

More information

Convergence of Rump s Method for Inverting Arbitrarily Ill-Conditioned Matrices

Convergence of Rump s Method for Inverting Arbitrarily Ill-Conditioned Matrices published in J. Comput. Appl. Math., 205(1):533 544, 2007 Convergence of Rump s Method for Inverting Arbitrarily Ill-Conditioned Matrices Shin ichi Oishi a,b Kunio Tanabe a Takeshi Ogita b,a Siegfried

More information

A priori bounds on the condition numbers in interior-point methods

A priori bounds on the condition numbers in interior-point methods A priori bounds on the condition numbers in interior-point methods Florian Jarre, Mathematisches Institut, Heinrich-Heine Universität Düsseldorf, Germany. Abstract Interior-point methods are known to be

More information

Some bounds for the spectral radius of the Hadamard product of matrices

Some bounds for the spectral radius of the Hadamard product of matrices Some bounds for the spectral radius of the Hadamard product of matrices Tin-Yau Tam Mathematics & Statistics Auburn University Georgia State University, May 28, 05 in honor of Prof. Jean H. Bevis Some

More information

Bounds on the radius of nonsingularity

Bounds on the radius of nonsingularity Towards bounds on the radius of nonsingularity, Milan Hladík Department of Applied Mathematics, Charles University, Prague and Institute of Computer Science of the Academy of Czech Republic SCAN 2014,

More information

Some preconditioners for systems of linear inequalities

Some preconditioners for systems of linear inequalities Some preconditioners for systems of linear inequalities Javier Peña Vera oshchina Negar Soheili June 0, 03 Abstract We show that a combination of two simple preprocessing steps would generally improve

More information

A NOTE ON CHECKING REGULARITY OF INTERVAL MATRICES

A NOTE ON CHECKING REGULARITY OF INTERVAL MATRICES A NOTE ON CHECKING REGULARITY OF INTERVAL MATRICES G. Rex and J. Rohn Abstract It is proved that two previously known sufficient conditions for regularity of interval matrices are equivalent in the sense

More information

The Solvability Conditions for the Inverse Eigenvalue Problem of Hermitian and Generalized Skew-Hamiltonian Matrices and Its Approximation

The Solvability Conditions for the Inverse Eigenvalue Problem of Hermitian and Generalized Skew-Hamiltonian Matrices and Its Approximation The Solvability Conditions for the Inverse Eigenvalue Problem of Hermitian and Generalized Skew-Hamiltonian Matrices and Its Approximation Zheng-jian Bai Abstract In this paper, we first consider the inverse

More information

Z-Pencils. November 20, Abstract

Z-Pencils. November 20, Abstract Z-Pencils J. J. McDonald D. D. Olesky H. Schneider M. J. Tsatsomeros P. van den Driessche November 20, 2006 Abstract The matrix pencil (A, B) = {tb A t C} is considered under the assumptions that A is

More information

Interlacing Inequalities for Totally Nonnegative Matrices

Interlacing Inequalities for Totally Nonnegative Matrices Interlacing Inequalities for Totally Nonnegative Matrices Chi-Kwong Li and Roy Mathias October 26, 2004 Dedicated to Professor T. Ando on the occasion of his 70th birthday. Abstract Suppose λ 1 λ n 0 are

More information

Notes on Linear Algebra and Matrix Theory

Notes on Linear Algebra and Matrix Theory Massimo Franceschet featuring Enrico Bozzo Scalar product The scalar product (a.k.a. dot product or inner product) of two real vectors x = (x 1,..., x n ) and y = (y 1,..., y n ) is not a vector but a

More information

Condition number and matrices

Condition number and matrices Condition number and matrices Felipe Bottega Diniz arxiv:70304547v [mathgm] 3 Mar 207 March 6, 207 Abstract It is well known the concept of the condition number κ(a) A A, where A is a n n real or complex

More information

Lecture 5. Ch. 5, Norms for vectors and matrices. Norms for vectors and matrices Why?

Lecture 5. Ch. 5, Norms for vectors and matrices. Norms for vectors and matrices Why? KTH ROYAL INSTITUTE OF TECHNOLOGY Norms for vectors and matrices Why? Lecture 5 Ch. 5, Norms for vectors and matrices Emil Björnson/Magnus Jansson/Mats Bengtsson April 27, 2016 Problem: Measure size of

More information

arxiv: v1 [math.na] 28 Jun 2013

arxiv: v1 [math.na] 28 Jun 2013 A New Operator and Method for Solving Interval Linear Equations Milan Hladík July 1, 2013 arxiv:13066739v1 [mathna] 28 Jun 2013 Abstract We deal with interval linear systems of equations We present a new

More information

The following two problems were posed by de Caen [4] (see also [6]):

The following two problems were posed by de Caen [4] (see also [6]): BINARY RANKS AND BINARY FACTORIZATIONS OF NONNEGATIVE INTEGER MATRICES JIN ZHONG Abstract A matrix is binary if each of its entries is either or The binary rank of a nonnegative integer matrix A is the

More information

Multiplicative Perturbation Analysis for QR Factorizations

Multiplicative Perturbation Analysis for QR Factorizations Multiplicative Perturbation Analysis for QR Factorizations Xiao-Wen Chang Ren-Cang Li Technical Report 011-01 http://www.uta.edu/math/preprint/ Multiplicative Perturbation Analysis for QR Factorizations

More information

Tikhonov Regularization of Large Symmetric Problems

Tikhonov Regularization of Large Symmetric Problems NUMERICAL LINEAR ALGEBRA WITH APPLICATIONS Numer. Linear Algebra Appl. 2000; 00:1 11 [Version: 2000/03/22 v1.0] Tihonov Regularization of Large Symmetric Problems D. Calvetti 1, L. Reichel 2 and A. Shuibi

More information

Nonlinear Programming Algorithms Handout

Nonlinear Programming Algorithms Handout Nonlinear Programming Algorithms Handout Michael C. Ferris Computer Sciences Department University of Wisconsin Madison, Wisconsin 5376 September 9 1 Eigenvalues The eigenvalues of a matrix A C n n are

More information

Error estimates for the ESPRIT algorithm

Error estimates for the ESPRIT algorithm Error estimates for the ESPRIT algorithm Daniel Potts Manfred Tasche Let z j := e f j j = 1,..., M) with f j [ ϕ, 0] + i [ π, π) and small ϕ 0 be distinct nodes. With complex coefficients c j 0, we consider

More information

Applied Mathematics Letters. Comparison theorems for a subclass of proper splittings of matrices

Applied Mathematics Letters. Comparison theorems for a subclass of proper splittings of matrices Applied Mathematics Letters 25 (202) 2339 2343 Contents lists available at SciVerse ScienceDirect Applied Mathematics Letters journal homepage: www.elsevier.com/locate/aml Comparison theorems for a subclass

More information

Matrix functions that preserve the strong Perron- Frobenius property

Matrix functions that preserve the strong Perron- Frobenius property Electronic Journal of Linear Algebra Volume 30 Volume 30 (2015) Article 18 2015 Matrix functions that preserve the strong Perron- Frobenius property Pietro Paparella University of Washington, pietrop@uw.edu

More information

Checking Properties of Interval Matrices 1

Checking Properties of Interval Matrices 1 Checking Properties of Interval Matrices 1 Jiří Rohn 2 1 The text, or a part of it, will appear in the book by Kreinovich, Lakeyev, Rohn and Kahl [24]. This work was supported by the Charles University

More information

Key words. conjugate gradients, normwise backward error, incremental norm estimation.

Key words. conjugate gradients, normwise backward error, incremental norm estimation. Proceedings of ALGORITMY 2016 pp. 323 332 ON ERROR ESTIMATION IN THE CONJUGATE GRADIENT METHOD: NORMWISE BACKWARD ERROR PETR TICHÝ Abstract. Using an idea of Duff and Vömel [BIT, 42 (2002), pp. 300 322

More information

On the loss of orthogonality in the Gram-Schmidt orthogonalization process

On the loss of orthogonality in the Gram-Schmidt orthogonalization process CERFACS Technical Report No. TR/PA/03/25 Luc Giraud Julien Langou Miroslav Rozložník On the loss of orthogonality in the Gram-Schmidt orthogonalization process Abstract. In this paper we study numerical

More information

Properties of Matrices and Operations on Matrices

Properties of Matrices and Operations on Matrices Properties of Matrices and Operations on Matrices A common data structure for statistical analysis is a rectangular array or matris. Rows represent individual observational units, or just observations,

More information

Introduction to Numerical Linear Algebra II

Introduction to Numerical Linear Algebra II Introduction to Numerical Linear Algebra II Petros Drineas These slides were prepared by Ilse Ipsen for the 2015 Gene Golub SIAM Summer School on RandNLA 1 / 49 Overview We will cover this material in

More information

Matrix Inequalities by Means of Block Matrices 1

Matrix Inequalities by Means of Block Matrices 1 Mathematical Inequalities & Applications, Vol. 4, No. 4, 200, pp. 48-490. Matrix Inequalities by Means of Block Matrices Fuzhen Zhang 2 Department of Math, Science and Technology Nova Southeastern University,

More information

the Unitary Polar Factor æ Ren-Cang Li Department of Mathematics University of California at Berkeley Berkeley, California 94720

the Unitary Polar Factor æ Ren-Cang Li Department of Mathematics University of California at Berkeley Berkeley, California 94720 Relative Perturbation Bounds for the Unitary Polar Factor Ren-Cang Li Department of Mathematics University of California at Berkeley Berkeley, California 9470 èli@math.berkeley.eduè July 5, 994 Computer

More information

THE RELATION BETWEEN THE QR AND LR ALGORITHMS

THE RELATION BETWEEN THE QR AND LR ALGORITHMS SIAM J. MATRIX ANAL. APPL. c 1998 Society for Industrial and Applied Mathematics Vol. 19, No. 2, pp. 551 555, April 1998 017 THE RELATION BETWEEN THE QR AND LR ALGORITHMS HONGGUO XU Abstract. For an Hermitian

More information

Wavelets and Linear Algebra

Wavelets and Linear Algebra Wavelets and Linear Algebra 4(1) (2017) 43-51 Wavelets and Linear Algebra Vali-e-Asr University of Rafsanan http://walavruacir Some results on the block numerical range Mostafa Zangiabadi a,, Hamid Reza

More information

Journal of Computational and Applied Mathematics

Journal of Computational and Applied Mathematics Journal of Computational and Applied Mathematics 236 (212) 3218 3227 Contents lists available at SciVerse ScienceDirect Journal of Computational and Applied Mathematics journal homepage: www.elsevier.com/locate/cam

More information

arxiv: v3 [math.ra] 22 Aug 2014

arxiv: v3 [math.ra] 22 Aug 2014 arxiv:1407.0331v3 [math.ra] 22 Aug 2014 Positivity of Partitioned Hermitian Matrices with Unitarily Invariant Norms Abstract Chi-Kwong Li a, Fuzhen Zhang b a Department of Mathematics, College of William

More information

DEN: Linear algebra numerical view (GEM: Gauss elimination method for reducing a full rank matrix to upper-triangular

DEN: Linear algebra numerical view (GEM: Gauss elimination method for reducing a full rank matrix to upper-triangular form) Given: matrix C = (c i,j ) n,m i,j=1 ODE and num math: Linear algebra (N) [lectures] c phabala 2016 DEN: Linear algebra numerical view (GEM: Gauss elimination method for reducing a full rank matrix

More information

OUTLINE 1. Introduction 1.1 Notation 1.2 Special matrices 2. Gaussian Elimination 2.1 Vector and matrix norms 2.2 Finite precision arithmetic 2.3 Fact

OUTLINE 1. Introduction 1.1 Notation 1.2 Special matrices 2. Gaussian Elimination 2.1 Vector and matrix norms 2.2 Finite precision arithmetic 2.3 Fact Computational Linear Algebra Course: (MATH: 6800, CSCI: 6800) Semester: Fall 1998 Instructors: { Joseph E. Flaherty, aherje@cs.rpi.edu { Franklin T. Luk, luk@cs.rpi.edu { Wesley Turner, turnerw@cs.rpi.edu

More information

Row Space and Column Space of a Matrix

Row Space and Column Space of a Matrix Row Space and Column Space of a Matrix 1/18 Summary: To a m n matrix A = (a ij ), we can naturally associate subspaces of K n and of K m, called the row space of A and the column space of A, respectively.

More information

A Note on Simple Nonzero Finite Generalized Singular Values

A Note on Simple Nonzero Finite Generalized Singular Values A Note on Simple Nonzero Finite Generalized Singular Values Wei Ma Zheng-Jian Bai December 21 212 Abstract In this paper we study the sensitivity and second order perturbation expansions of simple nonzero

More information

Abed Elhashash, Uriel G. Rothblum, and Daniel B. Szyld. Report August 2009

Abed Elhashash, Uriel G. Rothblum, and Daniel B. Szyld. Report August 2009 Paths of matrices with the strong Perron-Frobenius property converging to a given matrix with the Perron-Frobenius property Abed Elhashash, Uriel G. Rothblum, and Daniel B. Szyld Report 09-08-14 August

More information

Spectral inequalities and equalities involving products of matrices

Spectral inequalities and equalities involving products of matrices Spectral inequalities and equalities involving products of matrices Chi-Kwong Li 1 Department of Mathematics, College of William & Mary, Williamsburg, Virginia 23187 (ckli@math.wm.edu) Yiu-Tung Poon Department

More information

EECS 275 Matrix Computation

EECS 275 Matrix Computation EECS 275 Matrix Computation Ming-Hsuan Yang Electrical Engineering and Computer Science University of California at Merced Merced, CA 95344 http://faculty.ucmerced.edu/mhyang Lecture 16 1 / 21 Overview

More information

CONVERGENCE OF MULTISPLITTING METHOD FOR A SYMMETRIC POSITIVE DEFINITE MATRIX

CONVERGENCE OF MULTISPLITTING METHOD FOR A SYMMETRIC POSITIVE DEFINITE MATRIX J. Appl. Math. & Computing Vol. 182005 No. 1-2 pp. 59-72 CONVERGENCE OF MULTISPLITTING METHOD FOR A SYMMETRIC POSITIVE DEFINITE MATRIX JAE HEON YUN SEYOUNG OH AND EUN HEUI KIM Abstract. We study convergence

More information

Algebra C Numerical Linear Algebra Sample Exam Problems

Algebra C Numerical Linear Algebra Sample Exam Problems Algebra C Numerical Linear Algebra Sample Exam Problems Notation. Denote by V a finite-dimensional Hilbert space with inner product (, ) and corresponding norm. The abbreviation SPD is used for symmetric

More information

Nonnegative and spectral matrix theory Lecture notes

Nonnegative and spectral matrix theory Lecture notes Nonnegative and spectral matrix theory Lecture notes Dario Fasino, University of Udine (Italy) Lecture notes for the first part of the course Nonnegative and spectral matrix theory with applications to

More information

Computing the Norm A,1 is NP-Hard

Computing the Norm A,1 is NP-Hard Computing the Norm A,1 is NP-Hard Dedicated to Professor Svatopluk Poljak, in memoriam Jiří Rohn Abstract It is proved that computing the subordinate matrix norm A,1 is NP-hard. Even more, existence of

More information

ON ORTHOGONAL REDUCTION TO HESSENBERG FORM WITH SMALL BANDWIDTH

ON ORTHOGONAL REDUCTION TO HESSENBERG FORM WITH SMALL BANDWIDTH ON ORTHOGONAL REDUCTION TO HESSENBERG FORM WITH SMALL BANDWIDTH V. FABER, J. LIESEN, AND P. TICHÝ Abstract. Numerous algorithms in numerical linear algebra are based on the reduction of a given matrix

More information

MAT 610: Numerical Linear Algebra. James V. Lambers

MAT 610: Numerical Linear Algebra. James V. Lambers MAT 610: Numerical Linear Algebra James V Lambers January 16, 2017 2 Contents 1 Matrix Multiplication Problems 7 11 Introduction 7 111 Systems of Linear Equations 7 112 The Eigenvalue Problem 8 12 Basic

More information

FAST VERIFICATION OF SOLUTIONS OF MATRIX EQUATIONS

FAST VERIFICATION OF SOLUTIONS OF MATRIX EQUATIONS published in Numer. Math., 90(4):755 773, 2002 FAST VERIFICATION OF SOLUTIONS OF MATRIX EQUATIONS SHIN ICHI OISHI AND SIEGFRIED M. RUMP Abstract. In this paper, we are concerned with a matrix equation

More information

Let x be an approximate solution for Ax = b, e.g., obtained by Gaussian elimination. Let x denote the exact solution. Call. r := b A x.

Let x be an approximate solution for Ax = b, e.g., obtained by Gaussian elimination. Let x denote the exact solution. Call. r := b A x. ESTIMATION OF ERROR Let x be an approximate solution for Ax = b, e.g., obtained by Gaussian elimination. Let x denote the exact solution. Call the residual for x. Then r := b A x r = b A x = Ax A x = A

More information

KKT Conditions for Rank-Deficient Nonlinear Least-Square Problems with Rank-Deficient Nonlinear Constraints1

KKT Conditions for Rank-Deficient Nonlinear Least-Square Problems with Rank-Deficient Nonlinear Constraints1 JOURNAL OF OPTIMIZATION THEORY AND APPLICATIONS: Vol. 100, No. 1. pp. 145-160, JANUARY 1999 KKT Conditions for Rank-Deficient Nonlinear Least-Square Problems with Rank-Deficient Nonlinear Constraints1

More information

SIGN PATTERNS THAT REQUIRE OR ALLOW POWER-POSITIVITY. February 16, 2010

SIGN PATTERNS THAT REQUIRE OR ALLOW POWER-POSITIVITY. February 16, 2010 SIGN PATTERNS THAT REQUIRE OR ALLOW POWER-POSITIVITY MINERVA CATRAL, LESLIE HOGBEN, D. D. OLESKY, AND P. VAN DEN DRIESSCHE February 16, 2010 Abstract. A matrix A is power-positive if some positive integer

More information

MULTIPLICATIVE PERTURBATION ANALYSIS FOR QR FACTORIZATIONS. Xiao-Wen Chang. Ren-Cang Li. (Communicated by Wenyu Sun)

MULTIPLICATIVE PERTURBATION ANALYSIS FOR QR FACTORIZATIONS. Xiao-Wen Chang. Ren-Cang Li. (Communicated by Wenyu Sun) NUMERICAL ALGEBRA, doi:10.3934/naco.011.1.301 CONTROL AND OPTIMIZATION Volume 1, Number, June 011 pp. 301 316 MULTIPLICATIVE PERTURBATION ANALYSIS FOR QR FACTORIZATIONS Xiao-Wen Chang School of Computer

More information

FRIEDRICH-ALEXANDER-UNIVERSITÄT ERLANGEN-NÜRNBERG. Lehrstuhl für Informatik 10 (Systemsimulation)

FRIEDRICH-ALEXANDER-UNIVERSITÄT ERLANGEN-NÜRNBERG. Lehrstuhl für Informatik 10 (Systemsimulation) FRIEDRICH-ALEXANDER-UNIVERSITÄT ERLANGEN-NÜRNBERG INSTITUT FÜR INFORMATIK (MATHEMATISCHE MASCHINEN UND DATENVERARBEITUNG) Lehrstuhl für Informatik 0 (Systemsimulation) On a regularization technique for

More information

Review of Basic Concepts in Linear Algebra

Review of Basic Concepts in Linear Algebra Review of Basic Concepts in Linear Algebra Grady B Wright Department of Mathematics Boise State University September 7, 2017 Math 565 Linear Algebra Review September 7, 2017 1 / 40 Numerical Linear Algebra

More information

On the simultaneous diagonal stability of a pair of positive linear systems

On the simultaneous diagonal stability of a pair of positive linear systems On the simultaneous diagonal stability of a pair of positive linear systems Oliver Mason Hamilton Institute NUI Maynooth Ireland Robert Shorten Hamilton Institute NUI Maynooth Ireland Abstract In this

More information

Index. book 2009/5/27 page 121. (Page numbers set in bold type indicate the definition of an entry.)

Index. book 2009/5/27 page 121. (Page numbers set in bold type indicate the definition of an entry.) page 121 Index (Page numbers set in bold type indicate the definition of an entry.) A absolute error...26 componentwise...31 in subtraction...27 normwise...31 angle in least squares problem...98,99 approximation

More information

A NEW EFFECTIVE PRECONDITIONED METHOD FOR L-MATRICES

A NEW EFFECTIVE PRECONDITIONED METHOD FOR L-MATRICES Journal of Mathematical Sciences: Advances and Applications Volume, Number 2, 2008, Pages 3-322 A NEW EFFECTIVE PRECONDITIONED METHOD FOR L-MATRICES Department of Mathematics Taiyuan Normal University

More information

On the adjacency matrix of a block graph

On the adjacency matrix of a block graph On the adjacency matrix of a block graph R. B. Bapat Stat-Math Unit Indian Statistical Institute, Delhi 7-SJSS Marg, New Delhi 110 016, India. email: rbb@isid.ac.in Souvik Roy Economics and Planning Unit

More information

Invertibility and stability. Irreducibly diagonally dominant. Invertibility and stability, stronger result. Reducible matrices

Invertibility and stability. Irreducibly diagonally dominant. Invertibility and stability, stronger result. Reducible matrices Geršgorin circles Lecture 8: Outline Chapter 6 + Appendix D: Location and perturbation of eigenvalues Some other results on perturbed eigenvalue problems Chapter 8: Nonnegative matrices Geršgorin s Thm:

More information

ACCURACY AND STABILITY OF THE NULL SPACE METHOD FOR SOLVING THE EQUALITY CONSTRAINED LEAST SQUARES PROBLEM

ACCURACY AND STABILITY OF THE NULL SPACE METHOD FOR SOLVING THE EQUALITY CONSTRAINED LEAST SQUARES PROBLEM BIT 0006-3835/98/3901-0034 $15.00 1999, Vol. 39, No. 1, pp. 34 50 c Swets & Zeitlinger ACCURACY AND STABILITY OF THE NULL SPACE METHOD FOR SOLVING THE EQUALITY CONSTRAINED LEAST SQUARES PROBLEM ANTHONY

More information

Conditions for Robust Principal Component Analysis

Conditions for Robust Principal Component Analysis Rose-Hulman Undergraduate Mathematics Journal Volume 12 Issue 2 Article 9 Conditions for Robust Principal Component Analysis Michael Hornstein Stanford University, mdhornstein@gmail.com Follow this and

More information

RELATIVE PERTURBATION THEORY FOR DIAGONALLY DOMINANT MATRICES

RELATIVE PERTURBATION THEORY FOR DIAGONALLY DOMINANT MATRICES RELATIVE PERTURBATION THEORY FOR DIAGONALLY DOMINANT MATRICES MEGAN DAILEY, FROILÁN M. DOPICO, AND QIANG YE Abstract. In this paper, strong relative perturbation bounds are developed for a number of linear

More information

Moore-Penrose s inverse and solutions of linear systems

Moore-Penrose s inverse and solutions of linear systems Available online at www.worldscientificnews.com WSN 101 (2018) 246-252 EISSN 2392-2192 SHORT COMMUNICATION Moore-Penrose s inverse and solutions of linear systems J. López-Bonilla*, R. López-Vázquez, S.

More information

Sign Patterns with a Nest of Positive Principal Minors

Sign Patterns with a Nest of Positive Principal Minors Sign Patterns with a Nest of Positive Principal Minors D. D. Olesky 1 M. J. Tsatsomeros 2 P. van den Driessche 3 March 11, 2011 Abstract A matrix A M n (R) has a nest of positive principal minors if P

More information

Perturbation results for nearly uncoupled Markov. chains with applications to iterative methods. Jesse L. Barlow. December 9, 1992.

Perturbation results for nearly uncoupled Markov. chains with applications to iterative methods. Jesse L. Barlow. December 9, 1992. Perturbation results for nearly uncoupled Markov chains with applications to iterative methods Jesse L. Barlow December 9, 992 Abstract The standard perturbation theory for linear equations states that

More information

Foundations of Matrix Analysis

Foundations of Matrix Analysis 1 Foundations of Matrix Analysis In this chapter we recall the basic elements of linear algebra which will be employed in the remainder of the text For most of the proofs as well as for the details, the

More information

Key words. Strongly eventually nonnegative matrix, eventually nonnegative matrix, eventually r-cyclic matrix, Perron-Frobenius.

Key words. Strongly eventually nonnegative matrix, eventually nonnegative matrix, eventually r-cyclic matrix, Perron-Frobenius. May 7, DETERMINING WHETHER A MATRIX IS STRONGLY EVENTUALLY NONNEGATIVE LESLIE HOGBEN 3 5 6 7 8 9 Abstract. A matrix A can be tested to determine whether it is eventually positive by examination of its

More information

1. Introduction. We consider linear discrete ill-posed problems of the form

1. Introduction. We consider linear discrete ill-posed problems of the form AN EXTRAPOLATED TSVD METHOD FOR LINEAR DISCRETE ILL-POSED PROBLEMS WITH KRONECKER STRUCTURE A. BOUHAMIDI, K. JBILOU, L. REICHEL, AND H. SADOK Abstract. This paper describes a new numerical method for the

More information

Two Characterizations of Matrices with the Perron-Frobenius Property

Two Characterizations of Matrices with the Perron-Frobenius Property NUMERICAL LINEAR ALGEBRA WITH APPLICATIONS Numer. Linear Algebra Appl. 2009;??:1 6 [Version: 2002/09/18 v1.02] Two Characterizations of Matrices with the Perron-Frobenius Property Abed Elhashash and Daniel

More information

The Skew-Symmetric Ortho-Symmetric Solutions of the Matrix Equations A XA = D

The Skew-Symmetric Ortho-Symmetric Solutions of the Matrix Equations A XA = D International Journal of Algebra, Vol. 5, 2011, no. 30, 1489-1504 The Skew-Symmetric Ortho-Symmetric Solutions of the Matrix Equations A XA = D D. Krishnaswamy Department of Mathematics Annamalai University

More information